Text preview for : HP-Bench-Briefs-1982-01-02.pdf part of HP HP-Bench-Briefs-1982-01-02 HP Publikacje HP-Bench-Briefs-1982-01-02.pdf
Back to : HP-Bench-Briefs-1982-01-0 | Home
u
a
E N FROM HEWLETT-PACKARD
JANUARY-FEBRUARY 1982
Functional
Analyzing
B y Jim Bechtold, Editor
Introduction
Functional analyzing usually pro-
vides the most direct and efficient
means of isolating a problem in an
electronic instrument. As another
term for logical troubleshooting,
functional analyzing is more appro-
priate and descriptive of the process
used in everyday bench service. Func-
tional analyzing is antithetic to shot-
gun troubleshooting because func-
tional analyzing involves analyzing Figure 1. Mike Keenan, HP Technical Specialist, substitutes part of a good DVM into a faulty
the unit in logical, sequential steps in unit as a form of haif-splitting.
order to isolate the defective circuit,
rather than jumping around replac- showed how the step-by-step proce- was turned on, all annunciator lights
ing components at random until the dure could be used to isolate a problem were illuminated and the display
problem goes away. in the electrical charging system of an remained blank. The DVM would not
automobile - a rather simple exam- respond to any switch settings. Before
A bonus to this logical, sequential ple. This article shows how these Mike applied AC power, he removed
diagnosis is that when you do discover same basic procedures are applied in the covers and looked and smelled for
the cause of the problem, you acquire troubleshooting a complex electronic evidence of broken components or
a kernel of knowledge that is added to instrument - a microprocessor con- heat. Everything seemed OK so Mike
your bank of experiences. Then, if a trolled 6%-digit integrating digital connected the DVM power cord and
similar problem occursagain, you can voltmeter (see Figure 1). switched the unit on - in fact, he
draw upon your bank of past experi- switched it on and off several times.
ences and mentally bracket the prob- To document the procedure, I sat at The DVM repeatedly locked up at
lem area. Troubleshooting a n in- the bench with one of Hewlett- turn-on just as the customer had indi-
strument by the shotgun method will Packard's repair center technicians, cated. Milking the front panel re-
generally effect a repair, but you will Technical Specialist Mike Keenan, vealed only that all the lights worked
gain no knowledge of what caused the and observed him use the trou- (except for the display, which was
trouble since you have only treated bleshooting procedures we have been blank).
the symptom - not diagnosed its discussing on an HP 3455A DVM. In
cause. less than 15 minutes he isolated the
cause of a complete failure to a single IN THIS ISSUE
Functional Analyzing component. Functional Analyzing
Applied to a Real Life
e The last issue of Bench Briefs de-
scribed the theory of logical trou-
The Customer Complained
that the DVM Locked Up at
Turn-On
Battery Cross-Reference
1982 Customer Training
Calendar
The first clue of trouble was the cus- Service Notes
bleshooting (see accompanying box
for a thumbnail description), and tomer's complaint that when the unit
Part No. 5952-0108 WWW.HPARCHIVE.COM
Logical Troubleshooting
The goal of logical troubleshooting 3. Have you or anyone else seen the remaining parts of the circuit that
is to isolate the defective area in the problem before? Or, is the symptom have not yet been checkeduntil you
instrument in as few steps as possi- listed in the service manual? Many find the faulty portion of the
ble. Getting started involves ask- times there is a list of troubles and circuit.
ing three questions: probable causes in the manual's
troubleshooting/service section. In conjunction with half-splitting
you will want to use the informa-
The next logical step is to milk the tion funnel approach; i.e., deciding
1. What is the product supposed instrument for all symptoms avail-
to do? This may seem simple- what type of check to make that is
able. Use your eyes, ears and nose. most appropriate to the size of the
minded but it will help you select
t h e required troubleshooting trouble brackets at a given time.
Then you want to bracket the prob-
equipment. lem, which is simply a means of es-
tablishing broad limits around the Performing the checks in this fun-
areas to be tested. This usually in- nel sequence assures you wide cov-
2. D e the product do what it was
os volves the next step which is half- erage of trouble possibilities
designed to do? In other words, was splitting. initially, but with low precision.
the customer trying to make the Then, gradually, as you proceed to
instrument do something it was Half-splitting involves making localize the trouble, the checks be-
not designed to do? checks a t the midpoints of the come more precise.
Explanation. This DVM is electrically
The Manual Recommends a
The next step was to see if anyoneelse
had experienced the same problem. and physically divided into two major Second Working DVM for
We could either ask other techs in the sections (with two major printed cir- Effective Half-Splitting
area,or seeifthe problem waslistedin cuit boards). The inguard section, Both the inguard and outguard sec-
the service manual - it was listed. which consists of the measuring cir- tions have their own separate micro-
We had now completed several pre- cuitry, a controller, and power processor controllers that handshake
liminary steps in the logical trou- supplies; and the outguard section, back and forth. Both sections are in-
bleshooting process (see accompany- which consists of most logic circuits terconnected through a single cable
ing box). We had: and their power supplies. The out- that can be easily unplugged. The
guard circuits function as the internal manual troubleshooting procedure
- Verified the product was not doing main controller, HP-IB interface, and recommends using a test cable (avail-
what it is supposed to do. front panel interface of the instru- able from HP) and a second, func-
- Milked the insides and front panel ment (see Figure 2). tional 3455 DVM. The test cable is
for clues.
- Looked to see if the problem
was common enough t o be
documented.
The Service Manual's
Troubleshooting Section
Helps Us Bracket the
Problem
The manual's troubleshootingsection
told us to first verify that the raw
power supplies were working prop-
erly, Mike checked and the power
supplies were fine.
The next troubleshooting step told us
to half-split the instrument since
either of its two main sections could Figure 2. This complex DVM can be easily half-split where the inguard and wtguard Sections
hang up the turn-on sequence. are interconnected at pdnt 8.
2 BENCH BRIEFS JANUARY-FEBRUARY 1982 WWW.HPARCHiVE.COM
of course, would be
cally found the prob-
'agnosedits cause, he would
The inguard troubleshooting section lar schematic that showed the -+lo-
emel of knowledge to his
related that the turn-on fault could be volt reference was derived f o the
rm bank of experiences. The shotgun
method would only t r e a t the
symptom. In other words, the next
failure of this type that Mike came up
against, he still wouldn't know what
caused the problem.
We lifted a couple of diodes in the
cifcuit that could have been shorted,
and then umldered a transistor that
isolated a n of the circuit* No
luck. Then Mike realized that 1 volts
0
wm being drawn down to 0.6 volts,
which was almost ground. He looked
for and found Q34,a transistor that, if
shorted or leaky, would draw the 10-
volt reference to ground. Refer to Fig-
ure 6. When we lifted one leg on the
transistor, the voltage came right
digital voltmeter. The basic principle
was to isolate the one defective stage
in the circuit by confirming all of the
other stages were working properly.
We accomplished this through half-
splitting. And the same simple prin-
ciple will work in most any type of
circuit. Remember, that as you are
half-splitting and moving those
brackets ever closer to the fault, you
are learning the instrument and
picking up those kernels of knowl-
edge for future use.
a
Figure 4. A portion o the 1tl0-volt reference Circdtry. Note that th8 kt0 volts is used by the
f
A/D converter as well as input and auto-calibration switching circuits, all shown on other
schematics.
Mike replaced Q34 and turned the HP The Same Approach is Used
3455 on. The DVM did not lock up and to TroubleshootAny Circuit Figure 5. 0 3 waa Wabed full on and 034
3
appeared to function properly. Mike was shorted. Therefore, the 10-volt +VREF
pressed the self-test button and the The example we coveredconcentrated was drawn to near ground potential (0.6
DVM passed with no other problems. on isolating a problem in a complex volts).
Customer Service Seminars
Digital Troubleshooting Techniques
An intensive, short service-oriented course that Specializedtools and techniques to troubleshoot these
introduces modern technology to the technologies.
analog repairman. Workshop- four hours ofhands-on experimentswith
gates and troubleshooting tools.
When - March22-25
August 10-13
SECOND DAY
Where - Instrument Support Division 0 Logicsymbo10gy. P
690 E. Middlefield Rd. Positive/Nerrative 1 I
Mt. View, CA 94042 UndersWd3ng the i'mplication of logic schematics.
(415) 969-0880 Implementation of logic gates: AND,OR,NOR,
Coordinator - Debra Mazenko NAND, XOR, Wired-OR.
Cost - $350/student Decoders and their uses.
Comparators and their uses.
Flip-flops: R-S, J-K(standard and master-slave).
D,
FIRST DAY
Analog vs. digital.
Workshop-four hours ofhands-on experimentswith
decoders, comparators and flip-flops. Students will B
0 IC Technology: DCTL, RTL, DTL, CTL, TTL, ECL, also have an opportunity to use modern tools to
EECL, HTL, MOS, I2L. troubleshoot faults in a printed circuit assembly.
1 4 BENCH BRIEFS JANUARY-FEBRUARY 1982 WWW. HPARCHIVE.COM
THIRD DAY FOURTH DAY
containing flip-flops: ROMsPROM (masked, E and UV).
a hexidecimal.
1 8ring counters.
-,
synchronous and ripple),
Numberingsystems including binary, BCD, octal and
RAMs: bipolar and MOS (static and dynamic).
Typical failures and the troubleshooting difficulties
encountered with ROMs, PROMS and RAMs.
Typical memory addressing techniques.
Introduction to binary math including half and full Modern display technologies, their application and
adders. common failure modes.
0 Workshop- four hours of hands-on time building and Introduction to the ROM-controlled device with
debugging counter circuits. emphasis on methods used to fault isolate.
Workshop- four hoursofexperiments leading to the
building of a functioning strobed display device.
Logic Analyzers
A service-oriented course for calibration and The course objective is to teach front panel control opera-
repair techniciansthat teaches application, circuit tions, circuit theory, and learn the fundamental compo-
theory and calibration. nents used throughout the unit. Other areas covered are
the power supply, trigger recognition, data acquisition and
When - April 5-8 storage, and the display circuitry.
Where - Colorado Springs Division
1900 Garden of the Gods Road The student is guided through the three fundamental
Colorado Springs, CO 80907 areas of logic analyzers:
(303) 598-1900
Coordinator - Mike Fredeen - Recognizing a trigger
Cost - $400/student - Storing the data
- DisDlavinP the data
e Hewlett-Packard, Colorado Springs Division, is offering
Service Training Seminars to customers on most all
models of Logic Analyzers. The courses are directed to
calibration and repair technicians and will teach appli-
cation, circuit theory, calibration, and troubleshooting to
component-level repair. Attendees should have some
prior knowledge of logic and oscilloscope circuits. Toward the end ofthe course, the instructor summarizesby
discussing overall troubleshooting from symptom to re-
pair. The student is shown how to " m i l k the front panel to
learn how failures affect the instrument's behavior. From
the behavior patterns, the student learns how to isolate the
fault to a particular function within the instrument and
finally to the faulty component.
Ordering Information - All courses must be ordered
through your nearest HP Sales Office. If you desire addi-
tional information about a specific course, contact the
coordinator at the course location.
Battery Cross Reference
When selecting a replacement bat- performance of the product be re- therefore a dead instrument; and HP
tery for your HP product, you may tained. While some batteries used in replacement parts are either not in
notice that your manual will lisL only HP instruments are identical to those stock or will take a few days to obtain.
a n HP part number, even though it that can be purchased a t a local elec- In these cases it may be worthwhile to
appears that this part is actually a tronics distributor, many batteries see if a substitute battery will work in
standard battery. Service personnel are either specially manufactured to the circuit. Perhaps an HP battery
often ask why only HP part numbers HP specifications, not available over could be ordered and installed at some
are used. the counter, or mechanically altered later date.
to fit certain HP applications.
It is a form of quality control. By To help you in these situations, here's
listing only HP part numbers we are a list of batteries, their HP part num-
recommending that HP replacement However, there may be situations bers, and the manufacturer's name
parts be used to ensure that original when you have a dead battery and and part number.
List of Battery Manufacturers
Number Manufacturer Number Manufacturer
00029 Deac 05139 MacLeod & Handpol Inc.
0 394
1 Gates Energy Products 05296 RCA Corp. Electronic Comp.
01420 Burgess Inc. 05452 Globe Battery Div. Globe-Union Inc.
01447 Gould Inc. 05469 Duracell InternationalInc.
01921 RCA Corp. Solid State Div. 05980 Aglo La Pile S A R L
02000 Muirhead Inst. Inc. 05987 Saft
02369 Gulton Battery Corp. 06195 Varta Battery AG
02413 Clevite Corp. Burgess Batt. Div. 06450 Yuasa Denki KK
02967 General Electric 06508 Japan Storage Battery Co.
031 13 Eppley Lab Inc. 07230 Sonnenschein GMBH
03551 Castall Inc. 07245 Panasonic
03602 Sonotone GSD Marathon Battery 07371 Sanyo Electric Inc.
03941 Honeywell Inc. 08709 Panasonic
04135 Universal Products Inc. 08781 Catalyst Research Corp.
04304 Union Carbide Corp. 08891 Power Conversion Inc.
04781 Ray-0-Vac Div. ESB Inc. 09593 Electrochem Industries Inc.
04910 Union Supply Co. 09839 Plainview Electronics Corp.
MFR MFR
HP P/N NUMBER^ MFR P
N
I HP P/N NUMBER 0
0 YFR P/N
1420-0001 01420 US308 142@-@033 04304 MC126
1420-0001 01304 M762S 142a-aO33 05469 MTR126
1420-0001 01781 11109 1420-1034 04304 I4504
1420-0002 01420 HZ30wX 142@-0@35 01420 MY15
142O-Oll2 04304 IY-350 142@-@@35 04304 I4505
1420-0002 04701 n711 1420-DO36 01304 M416
1420-0003 01420 n210 1420-0037 04304 I4457
1420-0@03 04304 nllso 14 20 1
I31I a3941 H362366
142@-@@@3 05469 IM-13X 1 42a I3IO1 05469 W30265l
142o-aoo4 01420 MH6-3R 1 41 E
2 I1 I
I9 05469 MTR-132R
1420-0004 a5469 HRlC3R 14 2 0 I4 0 , OS469 I4lR-135-R
1420-@105 05469 MTR-233R-316469 1* I4 1 05-9 MRMR
l420-00@6 05469 WTUZ~M 14 a I4 2 05469 M316295 TIPE R M l R
142o-oao1 02369 nu-4200 zo-Pc20 MOD. 14 0 I4I4 a3-1 n
1420-0007 03602 WSPEC. PlW23191 14 0 I4I4 139 n
1421-0008 03602 WPIN 23957 14 a I4I